key: cord-0923857-k7bh8bf7
authors: Liu, Xiaoyan; Li, Zhe; Liu, Shuai; Sun, Jing; Chen, Zhanghua; Jiang, Min; Zhang, Qingling; Wei, Yinghua; Wang, Xin; Huang, Yi-You; Shi, Yinyi; Xu, Yanhui; Xian, Huifang; Bai, Fan; Ou, Changxing; Xiong, Bei; Lew, Andrew M.; Cui, Jun; Fang, Rongli; Huang, Hui; Zhao, Jincun; Hong, Xuechuan; Zhang, Yuxia; Zhou, Fuling; Luo, Hai-Bin
title: Potential therapeutic effects of dipyridamole in the severely ill patients with COVID-19
date: 2020-04-20
journal: Acta Pharm Sin B
DOI: 10.1016/j.apsb.2020.04.008
sha: ffe7b1544ed9e8e4b2206bdd3c474e9e19085991
doc_id: 923857
cord_uid: k7bh8bf7

Abstract Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection can cause acute respiratory distress syndrome, hypercoagulability, hypertension, and multiorgan dysfunction. Effective antivirals with safe clinical profile are urgently needed to improve the overall prognosis. In an analysis of a randomly collected cohort of 124 patients with Corona Virus Disease 2019 (COVID-19), we found that hypercoagulability as indicated by elevated concentrations of D-dimers was associated with disease severity. By virtual screening of a U.S. Food and Drug Administration (FDA) approved drug library, we identified an anticoagulation agent dipyridamole (DIP) in silico, which suppressed SARS-CoV-2 replication in vitro. In a proof-of-concept trial involving 31 patients with COVID-19, DIP supplementation was associated with significantly decreased concentrations of D-dimers (P<0.05), increased lymphocyte and platelet recovery in the circulation, and markedly improved clinical outcomes in comparison to the control patients. In particular, all 8 of the DIP-treated severely ill patients showed remarkable improvement: 7 patients (87.5%) achieved clinical cure and were discharged from the hospitals while the remaining 1 patient (12.5%) was in clinical remission.

As of April 3, 2020, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2, formerly known as 2019-nCoV) 1,2 had infected over 1,000,000 patients in 200 countries, such as USA, Spain, Italy, Germany, France, and UK; this rapid spread has been declared a global pandemic. To date, no agents have been reported to be specific to treat severely ill patents. Identification of readily available drugs for repositioning in Corona Virus Disease 2019 (COVID-19) therapy avails a relatively rapid way to clinical treatment 3 .

SARS-CoV-2, together with SARS-CoV and MERS-CoV, belongs to the beta-coronavirus genus, which is an enveloped, positive-stranded RNA virus with approximately 30,000 nucleotides 4, 5 .

Angiotensin I converting enzyme 2 (ACE2) is the receptor that engages the Spike surface glycoprotein of SARS-CoV and SARS-CoV-2 6, 7 . ACE2 is highly expressed in many organs, including the lung, heart, kidney, and intestine. Notably, in experimental models of SARS-CoV infection, Spike protein engagement decreases ACE2 expression and activates the renin-angiotensin system (RAS) 6 . RAS activation promotes platelet adhesion and aggregation, and increases the risk for pulmonary embolism, hypertension and fibrosis [8] [9] [10] [11] . It also accelerates cardiac and kidney injury by increasing local angiotensin II concentrations [12] [13] [14] . Apart from affecting the classic RAS pathway, ACE2 deficiency in the intestine is associated with malnutrition and colonic inflammation 15 .

Infection from SARS-CoV can result in severe lymphopenia, prolonged coagulation profiles, lethal acute respiratory distress syndrome (ARDS), watery diarrhea, cardiac disease, and sudden death 9, [16] [17] [18] . Many features have also been reported for COVID-19, such as prolonged coagulation profiles, elevated concentrations of D-dimers, severe lymphopenia, ARDS, hypertension, and acute heart injury in ICU-admitted patients 2, 19 . Given that angiotensin II concentrations were highly elevated in the SARS-CoV-2 infected patients 20 , RAS was likely a major pathogenic contributor of disease progression. Indeed, in a recent study describing 1099 patients with COVID-19, the concentrations of D-dimers were elevated in 40% and 60% of the non-severe and severe cases at hospital admission 21 , respectively. Furthermore, Zhou et al. 22 showed that a concentration of D-dimer greater than 1 mg/L on admission was associated with significantly increased risk of mortality for patients with COVID-19. Thus, prophylactic anti-coagulation therapy should be considered for alleviating the multi-organ damage for patients with COVID-19.

After viral entry to the host cells, the coronavirus messenger RNA is first translated to yield the polyproteins, which are subsequently cleaved by two viral proteinases, 3C-like protease (3CLP, aka nsp5 or Mpro) and papain-like protease (PLP, or nsp3), to yield non-structural proteins essential for viral replication 23 . Inhibitors that suppress the activity of these proteases may inhibit viral replication and offer an avenue for the SARS-CoV-2 therapy. Dipyridamole (DIP) is an antiplatelet agent and acts as a phosphodiesterase (PDE) inhibitor that increases intracellular cAMP/cGMP 24 . Apart from the well-known antiplatelet function, DIP may provide potential therapeutic benefits to patients with COVID-19. First, published studies 25-30 , including clinical trials conducted in China [31] [32] [33] , have demonstrated that DIP has a broad spectrum antiviral activity, particularly efficacious against the positive-stranded RNA viruses 26 . Second, it suppresses inflammation and promotes mucosal healing 34 . Third, as a pan-PDE inhibitor, DIP may prevent acute injury and progressive fibrosis of the lung, heart, liver, and kidney 35 . Here we provide evidence advocating DIP as an adjunctive therapy.

We virtually screened a U.S. Food and Drug Administration (FDA) approved drug library and found that DIP bound to the SARS-CoV-2 protease Mpro ( Fig. 1A and Supporting Information Fig. S1 ).

Hydrophobic and hydrogen bond (H-bond) interactions are the main driving forces for the binding between DIP and Mpro. By free energy perturbation calculations, the binding free energy of ∆G pred was -8.60 kcal/mol with a predicted IC 50, pred value of 490 nmol/L by the equation ∆G pred =−RT ln (IC 50, pred ). The inhibitory potency of DIP against Mpro was then subjected to an enzymatic assay using a previously published method 36 . As a result, DIP exhibited an IC 50, exp value of 530±10 nmol/L 4 ( Fig. 1B) , which was consistent with the theoretical prediction of the IC 50, pred values.

To directly demonstrate that DIP suppresses SARS-CoV-2 replication in vitro, we measured viral titers using a susceptible cell line, the Vero E6 cells. Chloroquine was used as a positive control 37, 38 .

Remarkably, at concentration 100 nmol/L, DIP suppressed more than 50% of SARS-CoV-2 replication (Fig. 1C ). This is four times less than the predicted and experimentally confirmed IC 50 to suppress Mpro activity, which is consistent with previous findings showing that DIP possesses additional antiviral effects [25] [26] [27] [28] [29] [30] . DIP (50 mg oral TID) has been used in patients to prevent hypercoagulability 39 , and the serum drug concentration was reported to be around 3 µmol/L 40 .

Collectively, these data suggest that the therapeutic dosages of DIP used to treat hypercoagulability could potentially suppress SARS-CoV-2 replication in the infected patients.

Insert Fig. 1 

We first retrospectively analyzed a randomly collected cohort of 124 patients with COVID-19. This has revealed that decreased lymphocyte counts, increased concentrations of D-dimers, CRP, and IL-6 concentrations were significantly associated with disease severity (Table 1) .

To evaluate the therapeutic potential of DIP as an adjunctive therapy to promote virus clearance Baseline characteristics of the two groups were shown in Table 2 . The average ages of the patients were 56 years. All patients manifested a cough, >75% had shortness of breath, and 35%-57% had nausea and vomiting. Chest CT scan revealed bilateral pneumonia in the 14 DIP-treated patients and 17 patients in the control group. In addition, RT-PCR test of SARS-CoV-2 RNA was positive for all patients. D-dimer concentrations were elevated in 50% (4/8) and 42% (5/12) of the severely ill patients in the DIP-treated group and the control group, respectively. Comorbidities, including diabetes mellitus, cardiovascular, and cerebrovascular diseases, were found in 6 patients in each of the DIP and control groups. Patients with diabetes (Table 2) were treated with insulin injections, and those with cardiovascular diseases were treated with nifedipine.

Insert Table 2 2

DIP adjunctive therapy was provided in 14 patients. The treatment protocol comprised of 50 mg oral tablets administered thrice daily (a total of 150 mg) for 14 consecutive days. All patients received ribavirin, glucocorticoids, and oxygen therapy, but none received antifungal treatment. Mechanical ventilation was required for all the critically ill patients from the DIP-treated (n = 2), and 1 each from the severely and critically ill patients in the control group (n = 2). Other treatment included antibiotics (42.9% vs. 58.8%) and intravenous immunoglobulin (14.3% vs. 23.5%).

DIP adjunctive therapy was associated with markedly improved clinical cure and remission rates in both the non-severe and severely ill patients (odds ratio 23.75, P = 0.06; Tables 3 and 4 ). In particular, for the 8 severely ill patients in the DIP-treated group, 7 patients (87.5%) achieved clinical cure and were discharged from the hospitals, and the remaining 1 patient (12.5%) was in clinical remission. In contrast, for the 12 severely ill patients in the control group, 4 patients (33.3%) were discharged, 2 patients (16.7%) were in remission, and 2 patients (16.7%) died, respectively.

It should be mentioned that due to the urgent situation and the lack of resources to perform viral RNA detection by the participating hospitals, we were unable to accurately determine the effects of DIP to viral clearance. However, according to the qualitative RT-PCR result of SARS-CoV-2 RNA provided by local Centers for Disease Control and Prevention, the average time for virus clearance was shortened by 1.6 days for the severe cases in the DIP-treated group in comparison to the control group.

In analysis of the laboratory indices, we observed continuously increased, albeit not statistically significant, counts of lymphocyte and platelet in patients receiving DIP treatment in comparison to the control patients (Fig. 2) . Given that lymphocytopenia and thrombocytopenia are markers of disease severity for patients with COVID-19 20 , immune recovery may contribute to infection resolution in DIP-treated patients. It should be noted that 50% and 42% of the severely ill patients from the DIP-treated and control group had increased baseline concentrations of D-dimer, respectively ( Table 2) . We calculated the dynamic changes for each patient in reference to their own baseline value, and found that D-dimer rose continuously in the control group, whereas they were 6 decreased in the DIP-treated group (Fig. 2) .

Insert Fig. 2 

We also examined two critically ill patients who received DIP adjunctive therapy. A 70-year-old man who had suffered from hypoxia and multiorgan dysfunction at hospital admission unfortunately died His D-dimer concentration initially increased as high as 15.72 mg/L two days after DIP treatment, but has gradually declined to 2.79 mg/L 4-5 days after DIP adjunctive therapy. This reinforces that high concentrations of D-dimer and low lymphocyte counts are associated with poor prognosis and suggest that DIP treatment should be initiated before the progression to a critical state (Fig. 3B ). Fig. 3 

All patients received chest CT scans and showed typical multiple patchy ground-glass shadows in the lungs before the treatment. For the DIP-treated patients, the lesions from all patients had a varied degree of absorption after treatment. In the control group, CT images in 1 of the 12 severely ill patients showed progression ( Fig. 4 and Supporting Information Table S1 ). Fig. 4 

Despite the enormous threat of SARS-CoV-2, no drugs have been claimed to be specific including the existing drugs used to treat other viruses. In reference to SARS-CoV-2 infection, we hypothesized that the SARS-CoV-2 Spike protein engagement may activate RAS in the lung 6,41 . This hypothesis was supported by published clinical characteristics and biochemical data of the severe and critically ill patients with COVID-19, who showed ARDS, hypertension, acute heart, kidney injury, and positive D-dimer results 2, 19, 20 . In searching for available anticoagulants, we focused on DIP because of its broad-spectrum antiviral, anti-inflammatory, and anti-fibrotic effects. Very importantly, we found that an EC 50 value of 100 nmol/L to suppress SARS-CoV-2 replication in vitro, indicating that the therapeutic dosage of DIP may potentiate effective antiviral responses in infected patients. These findings are in concordance with our clinical findings of the overall remarkable outcomes in the severely ill patients receiving two weeks of DIP adjunctive therapy. All 7 the 8 DIP-treated severely ill patients showed remarkable improvement after DIP treatment, with 87.5% discharged from the hospitals and a further 12.5% showing clinical remission. In contrast, for the 12 severely ill patients in the control group, only 33.3% were discharged and death occurred in 16 .7%.

In a recent publication describing 1099 patients with COVID-19, D-dimer concentrations were elevated in 40% and 60% of the non-severe and severe cases at hospital admission 21 . It has been reported that a D-dimer concentration greater than 1 mg/L on admission was associated with significantly increased risk of mortality for patients with COVID-19 22 . We found that DIP adjunctive treatment blunted the increase in D-dimer concentrations, and increased the counts of circulating platelets and leucocytes. High concentrations of D-dimers are closely correlated with pulmonary embolism 42 , vascular thrombosis, and renal dysfunction 43 . It is a crucial prognostic factor and is important to determine whether ICU-patients recover from severe infections 44, 45 . Thus, prophylactic anti-coagulation therapy with DIP should be considered in patients with COVID-19 to reduce the risk of hypercoagulability and multi-organ damage.

It should be mentioned that several factors have limited our ability to fully investigate the therapeutic effects of DIP adjunctive therapy, these include the small number of enrolled patients, the lack of resources to quantify viral replication, and the requirement to follow the treatment guidelines under the circumstances of SARS-CoV-2 outbreak. However, we advocate further trials for DIP adjunctive therapy for patients with COVID-19, particularly for those with early signs of elevated concentrations of D-dimer. DIP has been used world-wide to treat coagulopathy. Additionally, it also exerts anti-inflammatory and antiviral effects in experimental settings and clinical trials. The wide availability, safety, and affordability of DIP argue for further investigation into its therapeutic use in COVID-19, particularly as SARS-CoV-2 infection has been declared a global pandemic. Clinical trial (ChiCTR2000030055) was registered. Informed written consents were obtained from all patients. The condition of the patients was monitored daily by the attending physicians. Routine laboratory test of the coagulation variables and blood indexes were carried out before, during, and after the treatment. Clinical symptoms and laboratory data were independently validated by two independent investigators for assurance of data accuracy.

SARS-CoV-2 RNA from nasopharyngeal swabs were detected upon request of the charging physicians by the local Centers for Disease Control and Prevention (Wuhan, China). Only qualitative data were available for the patients. 

As of February 8, 2020, 124 confirmed COVID-19 cases had been identified from Zhongnan Hospital of Wuhan University (Table 1) 

Anticoagulant therapy was provided via oral DIP tablets. The daily treatment protocol comprised of 9 possible adverse events. All patients received antiviral (ribavirin, 0.5 g, Q12h), corticoid (methylprednisolone sodium succinate, 40 mg, QID), oxygen therapy, and nutritional support as necessary. Patients with diabetes were treated with insulin injections (Table 2) , and those with cardiovascular diseases were treated with nifedipine.

We virtually screened an U.S. FDA-approved drug database using the SARS-CoV-2 protease Mpro as a drug target. DIP (PubChem CID: 3108, Fig. 1A ) was identified as a lead drug. In order to obtain the binding pattern and calculate the binding free energy between DIP and Mpro, DIP was firstly docked onto Mpro by using Glide-SP method with the default parameters 47 , and the optimal binding pose ( Fig. S1 ) was further assessed by absolute binding free energy calculation with free energy perturbation 48 . The calculations were carried out in Gromacs 2019 49 , and the thermodynamic cycle and procedure was similar to that used by Matteo et al. 50 . In the calculation, the ligand electrostatic and van der Waals interactions were decoupled using a linear alchemical pathway with ∆λ = 0.10 for the van der Waals and ∆λ = 0.20 for electrostatic interactions. Restraints were added for keeping the relative position between receptor and ligand, which consist of one distance, two angles, and three dihedrals harmonic potentials with a force constant of 10 kcal/mol/Å 2 [rad 2 ]. The distance and angles for the restraints were determined by the values of the last 2 ns of the 4 ns preliminary MD simulations. In the FEP calculations, 4 ns simulations were performed for each window. The sampled ∆U in the simulations were fitted by Gaussian algorithms and the free energy estimates were obtained by using the Bennet acceptance ratio (BAR) method 51 .

The detailed methods of enzymatic assays of Mpro are shown in Supporting Information S1. 

Statistical analyses and graphics production were performed using R v3.5.3 (Foundation for Statistical Computing) 52 and GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA).

Categorical variables were described as frequencies or percentages, and continuous variables were shown as mean with standard deviation/error. Comparison for two independent groups was conducted using Student's t test (for normally distributed data) or Mann-Whitney test (for non-normally distributed data). Comparison for laboratory indices between the DIP-treatment and control groups during the treatment course was conducted using generalized mixed linear model.

Logistic regression was performed to identify factors associated with the clinical outcomes. P < 0.05 was considered statistically significant. Detailed descriptions of data comparison and statistical tests were specified in the figure legends. manuscript for scientific content. All authors approved the final version of the article.

The authors have no conflicts of interest to declare. Tables  Table 1 Clinical variables in 124 patients with COVID-19. 

Epidemiological and clinical characteristics of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China

First Case of 2019 Novel Coronavirus in the United States

Genomic characterization of the 2019 novel human-pathogenic coronavirus isolated from a patient with atypical pneumonia after visiting Wuhan

Genome composition and divergence of the novel coronavirus (2019-nCoV) originating in China

A crucial role of angiotensin converting enzyme 2 (ACE2) in SARS coronavirus-induced lung injury

Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding

The pulmonary renin-angiotensin system

Angiotensin-converting enzyme 2 protects from severe acute lung failure

Angiotensin II AT1 receptor antagonists inhibit platelet adhesion and aggregation by nitric oxide release

Platelet activation in acute pulmonary embolism

Deletion of angiotensin-converting enzyme 2 accelerates pressure overload-induced cardiac dysfunction by increasing local angiotensin II

Loss of angiotensin-converting enzyme-2 leads to the late development of angiotensin II-dependent glomerulosclerosis

Angiotensin-converting enzyme 2 is an essential regulator of heart function

ACE2 links amino acid malnutrition to microbial ecology and intestinal inflammation

SARS-coronavirus modulation of myocardial ACE2 expression and inflammation in patients with SARS

Severe acute respiratory syndrome

The severe acute respiratory syndrome

Clinical characteristics of 138 hospitalized patients with 12 2019 novel coronavirus-infected pneumonia in Wuhan

Clinical and biochemical indexes from 2019-nCoV infected patients linked to viral loads and lung injury

Clinical characteristics of coronavirus disease 2019 in China

Bin Cao. Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Antiviral drugs specific for coronaviruses in preclinical development

Anti-platelet therapy: phosphodiesterase inhibitors

Antiviral action of dipyridamole and its derivatives against influenza virus A

Dipyridamole reversibly inhibits mengovirus RNA replication

Inhibition of herpes simplex virus reactivation by dipyridamole

Dipyridamole potentiates the inhibition by 3'-azido-3'-deoxythymidine and other dideoxynucleosides of human immunodeficiency virus replication in monocyte-macrophages

Evaluation of dipyridamole efficacy as an agent for preventing acute respiratory viral diseases

Epidemiological trial of the prophylactic effectiveness of the interferon inducer dipyridamole with respect to influenza and acute respiratory diseases

Efficacy of dipyridamole in the treatment of 116 children with acute upper respiratory tract infections

Treatment of viral upper respiratory tract infection in children with dipyridamole

Clinical observation of 45 cases of upper respiratory tract infection treated with dipyridamole

Mucosal profiling of pediatric-onset colitis and IBD reveals common pathogenics and therapeutic pathways

cAMP and Epac in the regulation of tissue fibrosis

Structure-based drug design, virtual screening and high-throughput screening rapidly identify antiviral leads targeting COVID-19

Chloroquine is a potent inhibitor of SARS coronavirus infection and spread

Remdesivir and chloroquine effectively inhibit the recently emerged novel coronavirus (2019-nCoV) in vitro

Clinical pharmacokinetics of dipyridamole

Distribution of dipyridamole in blood components among post-stroke patients treated with extended release formulation

The discovery of angiotensin-converting enzyme 2 and its role in acute lung injury in mice

Over-testing for suspected pulmonary embolism in american emergency departments: the continuing epidemic

Renal function-adjusted D-dimer levels in critically ill patients with suspected thromboembolism

Plasma D-dimer as a novel biomarker for predicting poor outcomes in HBV-related decompensated cirrhosis

A favorable outcome of dengue hemorrhagic fever despite poor prognostic indices: a case report with a mix of classic and unusual clinical and laboratory features

General Office of National Health Committee. Diagnosis and treatment scheme of novel coronavirus-infected pneumonia (trial version 6)

Glide: a new approach for rapid, accurate docking and scoring. 2. Enrichment factors in database screening

Absolute binding free energy calculation and design of a subnanomolar inhibitor of phosphodiesterase-10

GROMACS: fast, flexible, and free

Accurate calculation of the absolute free energy of binding for drug molecules

Efficient estimation of free energy differences from Monte Carlo data

Age (yr)-mean±SD (range)

56-96) acid of SARS-CoV-2-no. (%)

Comorbidities Diabetes mellitus-no. (%)

We cordially acknowledge Tencent Cloud and National Supercomputing centers in Guangzhou, 

Normal range